As a group of researchers at UX studio we tend to seek new methods and ideas that we can share with each other and use to improve our work. This year we’ve made this study method more formal and organized 3-month-long workgroups. We set up topics of interest for each workgroup and 4-5 people joined each – focusing on the same area. Our group set a goal to learn more about features and user needs prioritization. We asked one main question: how might we learn more about new and exciting feature prioritization methods in a structured way?
- Why does feature prioritization matter?
- Why do we use product stages?
We’ve read amazing articles describing and categorizing these techniques. At some point we realized that none of them show how to pick techniques based on the current stage of your product. How can one decide which feature prioritization method to use depending on the current stage of a product life cycle?
Why does feature prioritization matter?
At the end of the day the ultimate goal of product development is to decide what features to spend time on, in what order – and what to exile to the uncharted shores of the backlog. During product development, you have constant forces that push or pull you and feature candidates in opposite directions. For example:
- New feature ideas from the team
- What competitors have
- Speed to market
- Engineering resources
- User perception
- Investor interests
- Business value
- and so on…
Let’s be honest, it is really challenging for a product manager to decide and find the right mix of features to push through design and development. Especially in early phases of the development, where resources are scarce. That’s when prioritization techniques come to help and provide a sharp machete to cut through this dense jungle.
Even though you’ll never be able to do a perfect job – the final verdict will come from your users after release – you can help the team minimize the risk of assumption-based decision making by spending more time with proper feature prioritization (and research).
Why do we use product stages?
When picking a feature prioritization technique, we don’t recommend a one size fits all approach. The weight of those pushing and pulling forces vary depending on which stage your product is currently in. Every time you are about to pick a product prioritization method, you need to understand and carefully analyze most of the factors that contribute to your choice.
To help with that, we came up with “the researchers’ point of view” and put the challenges and our recommendations on the product life cycle.
In each product life-cycle stage you’ll read about:
- goals & circumstances of that stage
- challenges with finding the right method
- methods we recommend – with pros and cons
Stages of the product life cycle
All products start with an idea or two. You want to check if those ideas are feasible, do the necessary discovery research to find signals and validate your assumptions without risking too much. We assume you’ve done discovery research, and the road to developing the MVP has just begun. You may want to focus on expected, vital features, as well as features that might be more important to your investors if you’re not bootstrapping.
Develop your idea to a level where it can be first introduced to the market.
Challenges of feature prioritization:
The majority of decisions are on a strategic level at this point, so internal feature prioritization methods will most likely be regular activities. There’s a good chance that the team might have already piled up market research reports, but they still have a lot of assumptions – you want signals that support your decision in a cost-effective way.
How might we get closer to the market and the users, while we need to move fast and rely on internal prioritization? How do we balance out internal and external feature prioritization techniques?
Methods we recommend:
User story mapping or Systemico
These methods focus on customer value to prioritize features early on. They rely on breaking up the customer journey into user goals and see what user stories we need to satisfy to meet those goals. Set up assumpted goals of your future customers, validate them with research and identify features that customers will seek. Since this is an internal backlog prioritization method, you’ll need other methods and activities to do the actual mapping, validation and reprioritization with the team.
- It already uses user stories which most of us use.
- Customer value is the main factor
- Easy to cluster user stories for each release and focus on bare minimum
- Without being validated by research it remains assumption-based
- It does not take other factors into consideration (e.g. resources needed)
QFD (Quality Function Deployment)
This is a well-documented framework that is built on the relationship between customer problems and feature ideas. It can be an ideal way to rank feature ideas higher if they assumptively satisfy more than one customer problem and/or if they satisfy heavier problems.
- We can weigh in customer problems
- Well-documented framework
- Relies a lot on the score system, which can be a source of errors
- Takes more time than other, simpler methods
It’s the simplest internal method for small products. You put your feature ideas into one (or more) of these buckets: Must-have, Should-have, Could-have and Won’t-have.
- Quick and easy method
- No criteria, entirely depends on people participating in prioritization
- It might be hard to prioritize between features within the bucket if you have many
Lean prioritization (2×2 matrix)
Besides well-setup techniques you can use a matrix of 2 relevant factors to prioritize methods with your team. For example Impact vs. Effort or Value vs. Risk. At this stage these methods are fine, and can give you enough signal to move on. The method can be used in other stages as well; you just have to define clearly what you mean by value, since it can be different at each stage. E.g. value for products in growth value can mean how a feature can help you attract new users, while in the maturity stage how it can help you keep existing users.
- Quick and easy methods
- Easily done with workshops within the team
- If the descriptions of the factors are too abstract that might lead to ambiguity during prioritization with the team
As you can see at early stages we recommend using feature prioritization methods that mainly use internal team efforts for quick decision making. It’s important to run evaluative research with your product concept regularly and use those signals to shape your product.
At this stage you’re ready with an MVP (or have a more improved product than MVP). It’s time to show your product to the first customers and early adopters on the market. Basic features are competing with each other, because resources are finite and you have to be very mindful of what to develop next. On the other hand, you can already shoot for real user feedback from your alpha / beta or live product users.
Goal: Realizing Product-Market Fit, acquisition of users, heavy user research, reiterating.
Challenges of feature prioritization:
This can be the most expensive stage of the product life-cycle for a company: there is no market share yet, the sales are low while the cost of development, research and marketing are extremely high. With limited resources to find the perfect market fit you need continuous research then iteration, and trying to understand what your customers really want.
Methods we recommend:
Internal methods from the Development phase can work here too. However, now that you can gain customer insights, go and try methods that use those inputs actively. Since at this stage building your rapport is important (marketing, branding, UX already on full throttle, growth engines are starting soon) consider methods that use more than 1-2 factors in prioritization.
It is a method that combines both team input and user signals and tries to set up a relative, weighted rank of the feature ideas. It uses a score system to prioritize features and uses more factors than previous methods. Very similar to KANO and a bit to the QFD framework.
- Uses multiple factors (cost, benefit, risk)
- Uses both functional (benefit) and dysfunctional (penalty) aspects of features
- Leaves bigger room for the PO/team in weighing
- Has to set up own weight system carefully and frequently review it
It has a lot of commonalities with the Jobs To Be Done framework. Opportunity scoring is a quantitative method that tries to find the relationship between user satisfaction and feature importance by letting customers give scores on feature ideas.
- Quick and Easy method
- Identifies important features that users are unhappy with
- Relies a lot on user scoring
- You need a considerable amount of users to make it significant
This method uses a simple score matrix where each feature (or theme) is ranked based on the added criteria and its weight. You can add as many criterias as you want and set the weight of them to your liking. You can take into account the business value of the items on your list.
- Quick and Easy
- You can select your own system of criteria
- Weighting can be subjective, needs lot of testing
- Score system might be limiting by quantifying input
It’s very similar to Scorecards. But while Scorecards use a total number to set the rank of each feature, in Theme screening you select a baseline feature/theme and check it against the selected criteria. You give either +1, -1 or 0 to them compared to the baseline.
- Quick and Easy
- You select your own criteria
- No weighting compared to baseline
Okay, so you successfully launched your product, and the growth stage comes into play. You want to focus on pulling in and activating as many new users as you can. As the demand for the product starts to increase, you may want to work on establishing the product’s position on the market. With the increase of sales and with the improved profit margins of this stage you have the chance to invest further in the product. Now it is time to focus on product improvement and handling the competition – while adding features that ensure the constantly increasing number of new users.
Goal: user acquisition, expanding reach and product improvement
Challenges of feature prioritization:
Now it is time for you to listen even more to your target audience, to check what excites them. If you want your product to have a long life, you also need to create a thorough roadmap and plan the release of newer and newer features or improvements. You may want to understand better what could be a delighter for the users, to boost their satisfaction with the product. You also need a better understanding of the potential differentiating features compared to what the competitors can offer, to be able to keep the momentum of your product. Here are a few tips which methods can help you with that.
Methods we recommend:
It is quantitative and purely user-based, so this is the right method if you would like to see which features on your backlog satisfy or delight the users. It is a standardized survey to explore what the basic expectations of the users are and where the absence of these features would lead to frustration. Read more about Kano in our other article.
- Well-documented, complete system
- It can help you to continuously develop delighters
- Helps you get a clear picture of how users would prioritize features
- Easy to understand, very visual way to share the results of the survey
- It requires more time to gather the responses to the survey as you need a large sample size
- You cannot test more than 10-12 features without making the questionnaire too long
Frequency vs. Reach
This technique uses the lean prioritization method with two main questions: how many people will use this feature and how often?
- Helps with features where focusing on the right user group is key
- It’s mainly assumption-based
- Uses only two, limiting factors
Additionally, other methods from previous product lifecycle stages:
You probably have all essential features delivered. You have stable returns, you have a good position on the market and revenue is flowing in. But if you don’t update or innovate the product constantly it reaches its saturation point. People get used to it, they start losing interest as new and innovative direct or indirect competitors appear. You need to focus on enhancing your existing features and find the right extension strategy.
Goal: Retention, intact revenue, combat churn, maintain the market share and extend the life of the product.
So, what could mimic the growth phase in small, if it’s possible at all? How might we give back the momentum to the product while keeping the revenue intact?
Methods we recommend:
A largely internal framework from Intercom that uses Reach, Impact, Confidence, and Effort. It’s a solid multi-factor method that operates mostly with weighted estimations and little data. Read more about RICE here.
- Multi-factor method (reach, impact, effort)
- Its own ‘Confidence score’ keeps the estimations in the team’s hands
- Lots of assumption and estimation needed
- Need to set up and test a lot your own weighting (‘Importance multipliers’)
Buy A Feature
This fun innovation game can give a new, interactive perspective to your feature prioritization methods. With its projective approach, Buy a feature gives your team and your customers money as the metaphor of value and overall cost which they can use to purchase the features with the set up price.
- Fun, projective game
- Gives a limiting factor to user requirements
- Customers are not product experts
- Everything is very subjective, dependent on the feature price tag
Another good feature prioritization technique we described earlier for getting feedback on where to enhance your product:
Eventually, the available market for your product will start to shrink. It is either because you’ve already reached most people you catered for, or users start to switch to newer products. When the market becomes too saturated, it’s hard to find space for further growth. You can either try to find ways to completely overhaul your product or have a retirement strategy (with finding a replacing product).
Goal: Understanding which scenario is driving the decline, attempting to resurrect or retire the product.
At this point the cycle is closed, there is not much to do. So-called next-generation products attempt to shift the product back into the growth phase which requires prioritization techniques again starting with stage 1.
Product Owners have an enormous collection of feature prioritization methods, but sometimes it’s hard to pick the one to use in the right situation. Besides the product life cycle, we need to consider the following:
- Quantitative-Qualitative – Are we looking for large-scale numbers to drive decisions or are we good with answers to ‘whys’?
- Internal-External – Do we involve only the core team or people outside the organization?
- Domain – How much data you have as input? What kind of data is it?
- Factors involved – How many factors do you need to consider? (user value, cost, revenue, resources and so on)
- Complexity – How complicated is the method to execute? How much time does it take to execute? How resourceful it is to do it frequently all over again?
- Combination – Do you need to combine it with other methods to make it work? (an internal innovation game + a data-driven Kano)
These methods won’t make the decisions instead of us, but we can treat them as signals that support us to reduce the chaos among our features. Happy prioritizing!